Journal Issue

IMF-BIS conference: More timely, reliable real estate data can play key role in financial stability

International Monetary Fund. External Relations Dept.
Published Date:
November 2003
  • ShareShare
Show Summary Details

Despite their importance in macroeconomic policymaking and financial stability analyses, high-quality data on real estate indicators are scarce in many countries, and internationally comparable data do not exist. As a first step toward creating reliable real estate indicators, the IMF’s Statistics Department and the Bank for International Settlements (BIS) hosted a two-day conference on October 27-28 for officials and analysts from central banks, statistical institutes, international institutions, academia, and the private sector. Russell Krueger and Subramanian S. Sriram report on the proceedings of the confer-ence—the first of its kind at the international level.

Developments in the real estate sector can have economywide repercussions, and the IMF should be concerned, Managing Director Horst Köhler said in opening the conference, because “one of the primary areas of the IMF’s mission—the safeguarding of the stability of the international financial system—must necessarily deal with the relationships between real estate activity and price cycles, and the stability of banking institutions and financial systems.”

The organization’s own work on financial soundness indicators, he added, has pointed to the need to improve real estate statistics and gather more data on “financial sector exposures and risks related to residential and commercial real estate and construction.”

Clearly, real estate prices, as prices of assets, can play a significant role in macroeconomic policy, Carol Carson (Director of the IMF’s Statistics Department) observed. “They are used as information variables in making monetary policy decisions,” she said, and “they also provide insights on possible balance sheet problems that the lending institutions may be facing. Hence, they provide indications of financial system stability.”

With respect to the banking system, Susan Wachter (Wharton School of the University of Pennsylvania) and Bradford Case (Board of Governors of the U.S. Federal Reserve System) cautioned that “poor information and inadequate analysis of real estate risk contribute to the vulnerability of the banking system. Banks and individual managers have few data on which to base careful analysis of real estate prices.” Indeed, interactions between real estate and financial sector activity can generate potential and actual systemic instability. Participants cited as examples the U.S. savings and loan crisis, the financial crises in Sweden and Japan in the early 1990s, and the widespread real estate market collapses and financial crises in southeast Asia in the late 1990s.

Users are “desperate” for timely and quality measures—as Keith Hall (Reserve Bank of Australia) observed—but the accurate compilation of real estate data remains a difficult task. Real estate properties are heterogeneous, transactions of individual properties are infrequent, and the detailed information needed is difficult and costly to obtain. Steps can be taken, however, to improve data and enhance comparability.

How to compile better indicators

The conference, in fact, was designed to exchange ideas and search for consensus recommendations on such topics as what these data should look like, how they should be compiled, which international agencies should be involved in gathering comparable data, and the extent to which international standards and harmonization are called for.

Desirable characteristics. According to Stephan Arthur (BIS), good real estate data should have six attributes: availability, representativeness, comparability, continuity, length, and frequency. Given priorities and large gaps in data, Arthur placed less stress on timeliness, but Boaz Boon (CapitaLand, Singapore) and David Fenwick (U.K. Office for National Statistics) disagreed, arguing that timeliness is also important.

Methodology. Developing countries that are now compiling their own real estate indices often ask which methodologies to apply and which practices are appropriate. The conference examined the pros and cons of several methodologies for constructing price and value indices for residential real estate, commercial real estate, and nonresidential properties.

There appeared to be overall support for compiling hedonic indices, which are quality-adjusted price indices, or hybrids of hedonic indices with some other methodologies. Marc Prud’Homme (Statistics Canada) said, “Hedonics is the only way to go.” But several participants voiced concerns about the cost of collecting the required data and the methodology’s complexity. Anne Laferrere (National Institute for Statistics and Economic Studies, France) disagreed on the complexity issue. France, she said, successfully compiled an aggregate real estate price index using the hedonic index method. And if cost is an issue, she and several others stressed, administrative or processing records for real estate deeds, taxes, or mortgages often provide quick, detailed, and cost-effective data for compilation purposes.

Aggregate indicators or a family of indicators? International organizations like the BIS and the European Central Bank (ECB), government agencies in several developing and industrial countries, and private sector companies in some countries, like Absa Group in South Africa, have been compiling and/or attempting to build indices at the regional and national levels to measure developments in real estate and/or asset markets. Arthur said “aggregate asset price indices could prove a welcome addition to the set of variables considered by policymakers from the perspective of both monetary and financial stability,” but the consensus was that a family of indicators was needed. These disaggregated indicators would also be useful for numerous other official and private purposes, such as tax administration, land allocation policy, bank provisioning policies, and investment decisions in the property sector.

International standards and harmonization.

The conference pointed to the clear need for international standards in the construction of real estate indices and in the harmonization of definitions, methodologies, and standards across countries. “Harmonization of data and methodology would facilitate international comparisons,” said Kathleen Stephansen (Credit Suisse First Boston). Estrella Domingo (Philippines National Statistical Coordination Board), among several others, indicated that “the conference had increased the level of knowledge on real estate indicators,” while Naseer Ahmad (State Bank of Pakistan) and others expressed the hope that a guide could be prepared on best practices.

Participants also stressed that international cooperation and joint public-private sector efforts are key to efforts to harmonize standards on data and methodology. Ivan Matalik (Czech National Bank) saw “a role for the IMF and the BIS and other international institutions such as Eurostat or the ECB” in developing these standards. Future conferences and a network or a discussion group could also help promote developments in international standards and harmonization in methodology.

Who should collect data? Both public and private sector organizations are involved in compiling real estate indicators, and practices differ widely between countries. Some participants, such as Wachter and Arthur, argued for putting data collection in a “public domain,” while Rupert Nabarro (Investment Property Databank) countered that “it is important that the private sector operate in the area of collecting real estate prices for commercial properties.” The consensus favored a role for both in compiling timely and accurate measures of real estate indicators and in developing a well-coordinated system for collecting and sharing data.

How should countries proceed?

For countries that would like to compile real estate price indictors, the conference provided a number of useful suggestions. Because data gathering is costly, Case, among several others, recommended exploiting existing data sources and selecting the methodology accordingly. For countries just starting this work, Donald Haurin (Ohio State University) urged them to first decide on the problem to be solved and then to formulate a clear strategy for a data collection system and the appropriate methodology. “Better get the price index right,” he said, “because the effect of peaks and troughs may be explained by data errors.” Thus, data collection and compilation methodology should be carefully matched to needs.

What’s next?

Resources are constrained, Carson acknowledged, but the IMF will continue to work on real estate indicators, and, in doing so, it will continue to collaborate with other international organizations. Paul Van den Bergh (BIS) agreed, adding that “the networking established here will be useful” to support continued country efforts to compile real estate indicators. To keep up the momentum, he recommended that guidance and technical assistance be provided to emerging markets that are aiming to build real estate indicators. This conference, he reiterated, is only the first step. Pressure to resolve real estate data needs will not go away, he said, because risk managers, as key players in helping sustain financial stability, will need detailed and timely data. Van den Bergh said he looked forward to continued collaboration between the IMF and the BIS in this and in other statistical areas.

Final conference papers will be published in a book, expected in late 2004, designed as a resource for analysts and statistical compilers.

Other Resources Citing This Publication